Question

我正在尝试使用BigQuery的Python API创建新表：

bigquery.tables().insert(
    projectId="xxxxxxxxxxxxxx",
    datasetId="xxxxxxxxxxxxxx", 
    body='{
        "tableReference": {
            "projectId":"xxxxxxxxxxxxxx", 
            "tableId":"xxxxxxxxxxxxxx", 
            "datasetId":"accesslog"},
        "schema": {
            "fields": [
                {"type":"STRING", "name":"ip"},
                {"type":"TIMESTAMP", "name":"ts"},
                {"type":"STRING", "name":"event"},
                {"type":"STRING", "name":"id"},
                {"type":"STRING","name":"sh"},
                {"type":"STRING", "name":"pub"},
                {"type":"STRING", "name":"context"},
                {"type":"STRING", "name":"brand"},
                {"type":"STRING", "name":"product"}
            ]
        }
    }'
).execute()

我得到的错误是：

(<class 'apiclient.errors.HttpError'>, <HttpError 400 when requesting https://www.googleapis.com/bigquery/v2/projects/xxxxxxxxxxxxxx/datasets/xxxxxxxxxxxxxx/tables?alt=json returned "Required parameter is missing">, <traceback object at 0x17e1c20>)

我认为所有必需的参数都包括在https://developers.google.com/resources/api-libraries/documentation/bigquery/v2/python/latest/bigquery_v2.tables.html#insert

中

缺少什么？

Answer 1

tables.insert唯一必需的参数是tableReference，其必须包含tableId，datasetId和projectId字段。我认为实际的问题可能是你传递JSON字符串时只能传递dict值。例如，以下代码用于创建表（请注意dataset_ref是将内容复制到命名参数的Python技巧）：

project_id = <my project>
dataset_id = <my dataset>
table_id = 'table_001'
dataset_ref = {'datasetId': dataset_id,
               'projectId': project_id}
table_ref = {'tableId': table_id,
             'datasetId': dataset_id,
             'projectId': project_id}
table = {'tableReference': table_ref}
table = bigquery.tables().insert(
    body=table, **dataset_ref).execute(http)

Answer 2

可能为时已晚，但问题是body参数必须是字典而不是字符串。

通过Python API V2在Google BigQuery中创建新表时缺少参数

2 个答案: